Classification of Variants across paired longitudinal samples in the PBTA Cohort

Set up

suppressPackageStartupMessages({
  library(tidyverse)
})

Directories and File Inputs/Outputs

# Detect the ".git" folder -- this will be in the project root directory
# Use this as the root directory to ensure proper sourcing of functions 
# no matter where this is called from
root_dir <- rprojroot::find_root(rprojroot::has_dir(".git"))
analysis_dir <- file.path(root_dir, "analyses", "tmb-vaf-longitudinal")
results_dir <- file.path(analysis_dir, "results")
input_dir <- file.path(analysis_dir, "input")
files_dir <- file.path(root_dir, "analyses", "sample-distribution-analysis", "results")

# Input files
pbta_file <- file.path(files_dir, "pbta.tsv") # file from add-sample-distribution module
genomic_paired_file <- file.path(files_dir, "genomic_assays_matched_time_points.tsv")
tmb_vaf_file <- file.path(results_dir, "tmb_vaf_genomic.tsv")
palette_file <- file.path(root_dir, "figures", "palettes", "oncoprint_color_palette.tsv")

# File path to plot directory
plots_dir <-
  file.path(analysis_dir, "plots")
if (!dir.exists(plots_dir)) {
  dir.create(plots_dir)
}

source(paste0(root_dir, "/figures/scripts/theme.R"))
source(paste0(analysis_dir, "/util/function-create-barplot.R"))

Read in data and process

# Vector to order timepoints
timepoints <- c("Diagnosis", "Progressive", "Recurrence", "xDeceased", "Second Malignancy", "Unavailable")

pbta_df <- readr::read_tsv(pbta_file, guess_max = 100000, show_col_types = FALSE) %>% 
  select(Kids_First_Participant_ID, Kids_First_Biospecimen_ID, cg_multiple, cg_id, cgGFAC, tumor_descriptor)

tmb_vaf_df <- readr::read_tsv(tmb_vaf_file, guess_max = 100000, show_col_types = FALSE) %>% 
  filter(!tmb >= 10) %>% 
  select(Kids_First_Biospecimen_ID, Variant_Classification, gene_protein, mutation_count,   region_size, tmb, VAF)

genomic_paired_df <- readr::read_tsv(genomic_paired_file, guess_max = 100000, show_col_types = FALSE) %>%
  left_join(pbta_df, by = c("Kids_First_Participant_ID")) %>% 
  left_join(tmb_vaf_df, by = c("Kids_First_Biospecimen_ID")) %>%
  filter(!is.na(tmb))

no_samples_with_tmb <- print(length(unique(genomic_paired_df$Kids_First_Participant_ID)))

[1] 116

# Attention as some bs specimen might not have TMB.
# If that happens, we will end up with samples lacking timepoints.
# Let's identify these samples and remove them for now.

df <- genomic_paired_df %>%
  select(Kids_First_Participant_ID, tumor_descriptor) %>% 
  unique() %>%
  arrange(Kids_First_Participant_ID, tumor_descriptor) %>%
  group_by(Kids_First_Participant_ID) %>%
  summarize(tumor_descriptor_sum = str_c(tumor_descriptor, collapse = ";")) %>%
  filter(!tumor_descriptor_sum %in% c("Diagnosis", "Progressive", "Recurrence", "Second Malignancy", "Unavailable", "Deceased", "Progressive;Progressive")) %>%
  left_join(genomic_paired_df, by = c("Kids_First_Participant_ID")) %>%
  filter(!cg_id == "NA") %>% 
  mutate(tumor_descriptor = case_when(grepl("Deceased", tumor_descriptor) ~ "xDeceased",
                               TRUE ~ tumor_descriptor),
         match_id = paste(tumor_descriptor, Kids_First_Participant_ID, sep = "_"),
         tumor_descriptor = factor(tumor_descriptor),
         tumor_descriptor = fct_relevel(tumor_descriptor, timepoints))

Warning: There was 1 warning in `mutate()`.
ℹ In argument: `tumor_descriptor = fct_relevel(tumor_descriptor, timepoints)`.
Caused by warning:
! 1 unknown level in `f`: Unavailable

no_samples <- print(length(unique(df$Kids_First_Participant_ID)))

[1] 107

# Let's count number of samples 
count_df <- df %>% 
  group_by(tumor_descriptor, cg_id, match_id, Variant_Classification) %>% 
  dplyr::count(cg_id) 

Define parameters for plots

# Read color palette
palette_df <- readr::read_tsv(palette_file, guess_max = 100000, show_col_types = FALSE) 

# Define and order palette
palette <- palette_df$hex_codes
names(palette) <- palette_df$color_names

Alterations per timepoint

# Define parameters for function
x_value <- count_df$tumor_descriptor
title <- paste("Variant types in PBTA cohort", sep = " ")

# Run function
fname <- paste0(plots_dir, "/", "Alteration_type_timepoints_barplots.pdf")
print(fname)

[1] "/Users/chronia/CHOP/GitHub/pbta-tumor-evolution/analyses/tmb-vaf-longitudinal/plots/Alteration_type_timepoints_barplots.pdf"

p <- create_stacked_barplot_variant(count_df = count_df, x = x_value, palette = palette, title = title)
pdf(file = fname, width = 6, height = 6)

print(p)
dev.off()

quartz_off_screen 
                2 

Alterations per timepoint in each cancer type

Alterations per timepoint in each cancer type and kids_id

cg_id_id <- as.character(unique(count_df$cg_id)) 
cg_id_id <- sort(cg_id_id, decreasing = FALSE)
cg_id_id

 [1] "Adamantinomatous Craniopharyngioma"       "Atypical Teratoid Rhabdoid Tumor"         "Chordoma"                                
 [4] "Choroid plexus carcinoma"                 "CNS Embryonal tumor"                      "Craniopharyngioma"                       
 [7] "Diffuse midline glioma"                   "Dysembryoplastic neuroepithelial tumor"   "Embryonal tumor with multilayer rosettes"
[10] "Ependymoma"                               "Ewing sarcoma"                            "Ganglioglioma"                           
[13] "Glial-neuronal tumor"                     "Hemangioblastoma"                         "High-grade glioma"                       
[16] "Low-grade glioma"                         "Malignant peripheral nerve sheath tumor"  "Medulloblastoma"                         
[19] "Meningioma"                               "Neuroblastoma"                            "Neurofibroma/Plexiform"                  
[22] "Pilocytic astrocytoma"                    "Schwannoma"                              

# Define parameters for function
x_value <- count_df$match_id
title <- paste("Variant types in PBTA cohort across cancer groups", sep = " ")

# Run function
fname <- paste0(plots_dir, "/", "Alteration_type_timepoints_cg_id_kids_barplots.pdf")
print(fname)

[1] "/Users/chronia/CHOP/GitHub/pbta-tumor-evolution/analyses/tmb-vaf-longitudinal/plots/Alteration_type_timepoints_cg_id_kids_barplots.pdf"

p <- create_stacked_barplot_variant_cg_id(count_df = count_df, x = x_value, palette = palette, title = title)
pdf(file = fname, width = 30, height = 30)

print(p)
dev.off()

quartz_off_screen 
                2 

Alterations per timepoint in each cancer type and timepoint model

tm_df_plot <- df %>%
  filter(!is.na(timepoints_models)) %>% 
  group_by(tumor_descriptor, cg_id, timepoints_models, match_id, Variant_Classification) %>% 
  dplyr::count(cg_id)

tm <- as.character(unique(tm_df_plot$timepoints_models))
tm <- sort(tm, decreasing = FALSE)
tm

 [1] "Dx-Dec"         "Dx-Pro"         "Dx-Pro-Dec"     "Dx-Pro-Rec"     "Dx-Pro-Rec-Dec" "Dx-Rec"         "Dx-Rec-Dec"     "Dx-SM"         
 [9] "Pro-Dec"        "Pro-Rec"        "Pro-Rec-Dec"    "Rec-Dec"        "Rec-SM"        

# Loop through variable
for (i in seq_along(tm)){
  print(i)
  df_sub <- tm_df_plot %>%
      filter(timepoints_models == tm[i])
  
  # Define parameters for function
  x_value <- df_sub$match_id
  title <- paste(tm[i])
  
  if (i %in% c(1,2,6)){
    rows <- 2
    }else{
      rows <- 1
      }
  
  # Run function
  p <- create_stacked_barplot_variant_cg_id(count_df = df_sub, x = x_value, palette = palette, title = title, rows = rows)
  
}

[1] 1
[1] 2
[1] 3
[1] 4
[1] 5
[1] 6
[1] 7
[1] 8
[1] 9
[1] 10
[1] 11
[1] 12
[1] 13

sessionInfo()

R version 4.2.3 (2023-03-15)
Platform: aarch64-apple-darwin20 (64-bit)
Running under: macOS Ventura 13.5.2

Matrix products: default
LAPACK: /Library/Frameworks/R.framework/Versions/4.2-arm64/Resources/lib/libRlapack.dylib

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

attached base packages:
[1] grid      stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
 [1] flextable_0.9.2 ggthemes_4.2.4  lubridate_1.9.2 forcats_1.0.0   stringr_1.5.0   dplyr_1.1.2     purrr_1.0.1     readr_2.1.4    
 [9] tidyr_1.3.0     tibble_3.2.1    ggplot2_3.4.2   tidyverse_2.0.0

loaded via a namespace (and not attached):
 [1] sass_0.4.7              bit64_4.0.5             vroom_1.6.3             jsonlite_1.8.7          carData_3.0-5          
 [6] bslib_0.5.0             shiny_1.7.4.1           askpass_1.2.0           fontLiberation_0.1.0    yaml_2.3.7             
[11] gdtools_0.3.3           pillar_1.9.0            backports_1.4.1         glue_1.6.2              uuid_1.1-0             
[16] digest_0.6.33           promises_1.2.0.1        ggsignif_0.6.4          colorspace_2.1-0        htmltools_0.5.5        
[21] httpuv_1.6.11           gfonts_0.2.0            fontBitstreamVera_0.1.1 pkgconfig_2.0.3         httpcode_0.3.0         
[26] broom_1.0.5             xtable_1.8-4            scales_1.2.1            later_1.3.1             officer_0.6.2          
[31] fontquiver_0.2.1        tzdb_0.4.0              openssl_2.1.0           timechange_0.2.0        generics_0.1.3         
[36] farver_2.1.1            car_3.1-2               ellipsis_0.3.2          ggpubr_0.6.0            cachem_1.0.8           
[41] withr_2.5.0             cli_3.6.1               magrittr_2.0.3          crayon_1.5.2            mime_0.12              
[46] evaluate_0.21           fansi_1.0.4             xml2_1.3.5              rstatix_0.7.2           textshaping_0.3.6      
[51] tools_4.2.3             data.table_1.14.8       hms_1.1.3               lifecycle_1.0.3         munsell_0.5.0          
[56] zip_2.3.0               compiler_4.2.3          jquerylib_0.1.4         systemfonts_1.0.4       rlang_1.1.1            
[61] rstudioapi_0.15.0       labeling_0.4.2          rmarkdown_2.23          gtable_0.3.3            abind_1.4-5            
[66] curl_5.0.2              R6_2.5.1                knitr_1.43              fastmap_1.1.1           bit_4.0.5              
[71] utf8_1.2.3              rprojroot_2.0.3         ragg_1.2.5              stringi_1.7.12          parallel_4.2.3         
[76] crul_1.4.0              Rcpp_1.0.11             vctrs_0.6.3             tidyselect_1.2.0        xfun_0.39